1 /*
2 * Copyright (c) 2021 Arm Limited.
3 *
4 * SPDX-License-Identifier: MIT
5 *
6 * Permission is hereby granted, free of charge, to any person obtaining a copy
7 * of this software and associated documentation files (the "Software"), to
8 * deal in the Software without restriction, including without limitation the
9 * rights to use, copy, modify, merge, publish, distribute, sublicense, and/or
10 * sell copies of the Software, and to permit persons to whom the Software is
11 * furnished to do so, subject to the following conditions:
12 *
13 * The above copyright notice and this permission notice shall be included in all
14 * copies or substantial portions of the Software.
15 *
16 * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
17 * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
18 * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
19 * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
20 * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
21 * OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
22 * SOFTWARE.
23 */
24 #ifndef SRC_CORE_NEON_KERNELS_CROP_CROP_HELPER_H
25 #define SRC_CORE_NEON_KERNELS_CROP_CROP_HELPER_H
26
27 #include "src/core/NEON/wrapper/wrapper.h"
28
29 namespace arm_compute
30 {
31 namespace cpu
32 {
33 template <typename T>
load_as_f32(T * ptr)34 inline float32x4_t load_as_f32(T *ptr)
35 {
36 ARM_COMPUTE_UNUSED(ptr);
37 ARM_COMPUTE_ERROR("Type not supported.");
38 }
39
40 #if defined(__ARM_FEATURE_FP16_VECTOR_ARITHMETIC) && defined(ENABLE_FP16_KERNELS)
41 template <>
load_as_f32(float16_t * ptr)42 inline float32x4_t load_as_f32(float16_t *ptr)
43 {
44 return vcvt_f32_f16(wrapper::vload(ptr));
45 }
46 #endif /* defined(__ARM_FEATURE_FP16_VECTOR_ARITHMETIC) && defined(ENABLE_FP16_KERNELS) */
47
48 template <>
load_as_f32(float * ptr)49 inline float32x4_t load_as_f32(float *ptr)
50 {
51 return wrapper::vloadq(ptr);
52 }
53
54 template <>
load_as_f32(int32_t * ptr)55 inline float32x4_t load_as_f32(int32_t *ptr)
56 {
57 return vcvtq_f32_s32(wrapper::vloadq(ptr));
58 }
59
60 template <>
load_as_f32(uint32_t * ptr)61 inline float32x4_t load_as_f32(uint32_t *ptr)
62 {
63 return vcvtq_f32_u32(wrapper::vloadq(ptr));
64 }
65
66 template <>
load_as_f32(int16_t * ptr)67 inline float32x4_t load_as_f32(int16_t *ptr)
68 {
69 return vcvtq_f32_s32(vmovl_s16(wrapper::vload(ptr)));
70 }
71
72 template <>
load_as_f32(uint16_t * ptr)73 inline float32x4_t load_as_f32(uint16_t *ptr)
74 {
75 return vcvtq_f32_u32(vmovl_u16(wrapper::vload(ptr)));
76 }
77
78 template <>
load_as_f32(uint8_t * ptr)79 inline float32x4_t load_as_f32(uint8_t *ptr)
80 {
81 return vcvtq_f32_u32(vmovl_u16(vget_low_u16(vmovl_u8(wrapper::vload(ptr)))));
82 }
83 }
84 } // namespace arm_compute
85
86 #endif //SRC_CORE_NEON_KERNELS_CROP_CROP_HELPER_H